DESCRIPTION Abstract: With the completion of the sequencing and annotation of the human genome, we now have a complete catalog of all human proteins encoded in the genome. However, the functions of a majority of the proteins remain unknown. One way to elucidate the functions of proteins is to find small molecule ligands that specifically bind to the proteins of interest and perturb their biochemical and cellular functions. A major challenge for chemical biologists today is to discover new small molecule probes for new proteins to facilitate the elucidation of their functions. The recent advance in the development of protein chips has offered an exciting new opportunity to simultaneously screen chemical libraries against nearly the entire human proteome for the first time. However, screening of the human protein chips is not yet feasible with most, if not all, existing chemical libraries due to the lack of a universal readout for detecting the binding of a ligand to a protein on chips. I propose to borrow a well-established protein-binding domain of natural products to serve both as a scaffold to present combinatorial peptide and non-peptide libraries and as an embedded universal tag for each compound in the library. Such hybrid combinatorial libraries will be amenable to proteome-wide screening using protein chips by exploiting the interaction between the protein binding domain derived from natural products and its binding protein, which can be detected with a fluorescently labeled antibody against the protein. In addition, known protein-binding domain can also confer stability and cell permeability to the fused ligands, increasing the probability that hits from such hybrid combinatorial libraries will be readily applicable to study the cellular functions of the relevant target proteins. This work, if successful, may lead to a new structural class of ligands that can be used as probes of protein function.